Friday October 28th 2022, 19:13 - October 29th 2022 00:37 UTC
Friday November 11th 2022, 12:20 - 15:11 UTC
Customers using legacy endpoints to fetch concurrency limits and usage observed an elevated number of 504 errors.
Affected endpoints:
This increased error rate was caused by connection issues between our ingress routing service and our API gateway service due to an incorrectly configured timeout setting. This timeout setting was introduced during a migration to the new API gateway.
The team migrated back to the old API gateway to restore service, and we saw the error rate subside.
The team reviewed all settings to ensure everything would work as expected as they resumed and finished the migration.